Ulrich Schimschal and J. Wright
Bureau of Reclamation, Technical Services Center, Denver,
Colorado
During November 1994 Bureau of Reclamation personnel completed a borehole geophysical logging survey in 17 wells at the Butz Landfill Superfund Site, Jackson Township, Monroe County, Pennsylvania. The purpose of the survey was to (1) describe material in the wells (the lithology), (2) determine the frequency, aperture, and orientation of any fractures in these materials, (3) determine the porosity of the materials, and (4) evaluate lithologic units for correlation between wells.
Borehole televiewer (BHTV), caliper (CALI), sonic transit time (DT), fluid temperature (FTEMP), and sonic variable density (SONIC VD) logs were evaluated as fracture indicators. Borehole televiewer (BHTV) data were evaluated by the Borehole Geophysical Laboratory at Stanford University to determine dip, strike, and aperture of any fractures intersecting the wells.
Shale percentages and corrected density porosities were computed for each well. Natural gamma ray (GR) and neutron (NEUT) logs were used as the primary indicators of lithologic units for correlation between wells. Due to the environment of deposition, subsequent tilting of the strata, and wide spacing of the wells, correlation of specific stratigraphic units was arduous. By splicing pieces of gamma logs from different wells, a typical gamma ray correlalation log was constructed and used to correlate large units(Units 1,2,and 3) within the bedrock. Smaller subunit markers (a through J) were then correlated to provided better ties between wells.
Other logs used include density (DENS), induction resistivity (REST), spontaneous potential (SP), single point resistance (SPR), and sonic amplitude (AMPL).
Calculation of formation water resistivities provides "target zones" of possible liquid-phase Trichloroethane (TCE) contamination. These zones are highlighted and potential packer-seat locations are identified for future testing. Calculated permeability and hydraulic transmissivity provides an estimate of aquifer performance and possible contamination movement.
This report describes the field method and processing steps used to complete a borehole geophysical logging survey at the Butz Landfill Site. The report evaluates the results of the geophysical investigation and relates the geophysical interpretation to the geologic conditions of the subsurface at the site.
The purpose of the investigation was to (1) describe the lithology of rock units on seventeen selected wells, (2) determine the fracture frequency, aperture, and orientation of any fractures in these materials, (3) determine formation porosity , and (4) evaluate lithologic units for correlation between wells. Fieldwork and preliminary data processing were done from November 1 through November 16, 1994.
Butz Landfill is in Jackson Township, Monroe County, Pennsylvania, about 6 miles west of the community of Tannersville. The site is along the western edge of the glaciated Low Plateaus section of the Pocono Plateau, within the Appalachian Plateau Physiographic Province. The bedrock at the site consists of alternating fine-grained sandstone and red shale and lesser amounts of siltstone and mudstone. The rocks probably were deposited in a lower delta plain. The bedrock was later folded and faulted.
Borehole geophysical logs were obtained in 17 wells at the Butz Landfill site during November, 1994. Well numbers, and locations are given in Figure 1. The primary logs obtained in the survey included natural gamma ray (GR), neutron (NEUT), density (DENS), caliper (CALI), induction resistivity (REST), spontaneous potential (SP), single point resistance (SPR), fluid temperature ( FTEMP), sonic transit time (DT), sonic amplitude (AMPL), sonic variable density (SONIC VD), and borehole televiewer (BHTV).
The geophysical logs obtained from the Butz Landfill Site were recorded and stored digitally using the ACQUIRE software purchased from COLOG Inc. The digital files were transferred into a log-analysis package, QLA2 Schlumberger, purchased from Geographix Inc. Processing of the data included minor depth shifting of traces, picking of compressional times, compensated density corrections, estimation of shale percentages, estimation of formation porosities, digitization and picking of borehole televiewer data, location of "target zones' for possible organics contamination, estimation of permeability and transmissivity.
Compressional wave velocities were calculated using a zero-crossing threshold method from the ACQUIRE software package. This method picks the first acoustic arrival on each receiver after a zero crossing and a set threshold are exceeded on each waveform.
Material density values were obtained using a dual- detector density probe and applying a spine-and-rib correction for probe standoff. Schimschal (1993) provides a detailed description of the density correction procedure.
Borehole televiewer data were digitized and interpreted under contract to Reclamation by Dr. Colleen Barton of Stanford University, Department of Geophysics, Borehole Geophysics Laboratory. Digitization of the standard analog data results in the dual measurement of the acoustic reflectivity of the borehole wall and the ultrasonic travel time of the imaging pulse. Included on the data log is a merged magnetic north pulse for image orientation. The analysis program utilizes various two- and three-dimensional displays of the borehole radius and acoustic reflectivity and allows the user to quantitatively measure borehole features such as breakouts, fracture orientations, apparent fracture aperture, and primary lithologic features. Barton et al (1991) provide a detailed description of the BHTV acquisition and analysis procedure.
Shale percentages and formation porosities were computed using a commercially available software package that has a statistical solution. Commercially available programs permit estimates of the relative proportions of mineral or lithologic components and pore fractions in the lithologic units determined in the logs from a borehole. Bed thickness resolution is approximately 0.5 feet.
The logs used must allow the differentiation of lithologic components. Ideally ,a log should be available for each lithologic component to be resolved. These lithologic components are calculated as fractions of the total rock matrix, including porosity.
When clay or shale are included in these calculations, the resulting porosity can be considered as effective matrix porosity. The interpreter must provide suitable parameters for each log and each component from either the available literature or suitable picks on the logs. As a check on the suitability of these parameters, the computer program allows the recalculation of the original logs in the logging suite from the mineral or lithologic components assumed to be present.
A comparison between these recalculated logs and the original field logs permits the interpreter to check the validity of the assumptions. Whenever possible data from mud logs or cores should be used as aid to adjust the mathematical results. In a borehole that has many different components, this approach may be the only one to obtain representative porosities. When viewing the lithologic display column, keep in mind that the individual components plus porosity add up to 100 percent across the depicted scale.
Stratigraphic units were correlated by overlaying the natural gamma logs from each well and visually comparing curve signatures. Due to wide separation of the wells and the moderately dipping strata in the area, any individual well
provides only limited coverage of the stratigraphic section beneath the Butz Landfill Site. Therefore, a "correlation log" was generated by combining natural gamma logs from several wells. The gamma logs from each well were then "tied" to the correlation log, and three major unit boundaries plus ten subunit markers were established. The major stratigraphic units were defined as follows:
Unit 1 (Unit 1-A, Unit 1-B, Unit 1-C, Unit 1-D, Unit 1-E). Alternating silty sandstones and shales characterized by well developed, fining-upward cyclic sequences.
Unit 2 (Unit 2-F, Unit 2-G, Unit 2-H). Alternating silty sandstones and shales characterized by poorly developed cyclic sequences composed mostly of shale and siltstone.
Unit 3 (Unit 3-I, Unit 3-J). Alternating silty sandstones and shales characterized by well developed fining-upward cyclic sequences composed mostly of massive-bedded silty sandstone.
Concentrations of organic compound were calculated using a statistical solution available in the PETRA module within the T-LOG log analysis package from Terrascience. An apparent formation-fluid resistivity is calculated from Archie's law for 100-percent water saturation. The Archie equation follows
Where:
Permeabilities were calculated from and experimentally established relationship. The equation to determine permeability (K) from porosity (N) follows:
Constants a and b were developed from injection pump test correlated with geophysical well log data. Permeabilities were calculated from an experimentally established relationship. The equation to determine permeability (K) from porosity ( f ) follows logs. Constants c and d reflect the effects of the specific logging tool, borehole environment, and lithology and are determined from cross plots (Schimschal 1981). The above equation was developed primarily for fractured rock and is thought to be applicable at the Butz site. Site specific experimentation might refine the constants a and b. The results from equation (2) are a good starting point for evaluation of the groundwater flow at the site.
Transmissivities were calculated from the permeabilities. The transmissivity at a given depth represent total groundwater flow between the water table and the particular depth on the log. For wells where the permeabilities were not calculated up to the water table due to steel casing transmissivities were extrapolated to the water table. The units of transmissivity were calculated in cubic feet per year considering the thickness between the water table and the given depth as the thickness of the aquifer, i.e. transmissivity (T) is given by:
Where K is permeability in cubic feet/ square foot-year and M is the aquifer thickness in feet.
An
example of a plot of the borehole geophysical logs obtained from
seventeen selected wells at the Butz Landfill Site is shown on Figure
3. Each plot contains eight tracks of data; CORRELATE, DENSITY,
ELECTRIC, FLUID, SONIC, LITHOPOR, SONIC VD, and INTERP BHTV.
Following is a discussion of the data presented in each track:
CORRELATE track includes natural gamma ray (GR) and neutron (NEUT) logs. Both logs are displayed in counts per second and are scaled to display variations in the lithology of the materials. Increases on the GR trace indicate increasingly finer material.
DENSITY track includes caliper (CALI) and compensated density (DENS)logs. CALI is displayed as hole diameter, in inches. Measurements of more than 6 inches are shaded. DENS is displayed in grams per cubic centimeter and is a compensated density computed from the dual-detector density probe as described earlier.
ELECTRIC track includes resistivity (REST) and Single-point resistance (SPR) logs. REST is displayed in ohmmeters and is computed from the induction conductivity probe. SPR, displayed in ohms, is a contact single-point measurement and is used for stratigraphic correlation only.
FLUID track includes spontaneous potential (SP), Fluid temperature (FTEMP), and delta temperature (DTEMP3) logs. SP is displayed in millivolts and is used to calculate formation-fluid resistivities. FTEMP is displayed in degrees Celsius and is generally run to correct resistivity measurements and to detect water movement. Because the temperature log was run after pulling the pumps in each well, the probes ability to detect water movement may be reduced. DTEMP3 is a three-point estimate of the derivative of the FTEMP curve to highlight small changes in the fluid temperature. The three-position forward-looking estimate (DTEMP3) was computed using the following equation: 1/(2h)[-y2+4*y1-3*y0], where the three y -values are y0, y1, y2 estimating the derivative at y0 and h is the sampling interval.
SONIC track includes sonic transit time unfiltered (dTunfilt), sonic transit time filtered (DT), and sonic wave amplitude (AMPL) logs. DT and dTunfilt are displayed in microseconds per foot. The DT trace was filtered using a weighted five-point filter to reduce the affects of cycle skipping. The two traces (DT and dTunfilt) were over plotted and the area between the curves shaded to accentuate zones of cycle skipping related to fractures within lithologic units.
LITHO POR track includes computed traces of percent shale, sandstone, and saturated porosity. Shale percentage was determined from a corrected gamma ray log and maximum and minimum counts for 100-percent shale and sandstone. Porosity was determined from a mathematical matrix solution incorporating the shale percent, formation density, matrix density, and formation-fluid density.
SONIC VD track includes the sonic variable density (SVD) log. The SVD provides a Z-axis display of the received sonic waveform from a single receiver. Irregularities along the waveform arrivals are generally associated with fractures and/or weathering of the material.
INTERP BHTV track includes tracings of fractures and bedding planes interpreted from the borehole televiewer data. The tracings are plotted from 0 to 360 degrees azimuth along the wall of the well. Planar features that intersect the borehole appear on the unwrapped 360 degree view as sinusoids. The amplitude and phase of the sinusoid are used to determine the dip and dip direction of the planar feature.
Figure 2 shows the stratigraphic units and subunit markers used to correlate the stratigraphy within the Butz Landfill Site. As described earlier, a correlation log was constructed from several natural gamma logs from different wells and was used to correlate the units and subunit markers present in individual wells.
After
units and subunit markers were correlated in each well, seven
geologic sections were constructed.. A plan map showing these section
locations is found on Figure 2. The sections were generally located
either along assumed strike (Sections 1,2,3,and 4) or perpendicular
to assumed strike (Sections 5,6, and 7). A plot of section 1-1 is
shown on Figure 2. Because these sections are hinged-point as opposed
to straight-line projected, the apparent dip of the strata changes
with the orientation of the bent section line. Shown on each section
are the well locations, subunit markers, and the natural gamma log
from each well. When a subunit marker is not present in an adjacent
well, the markers are projected along apparent dip with a dashed
line.
The presence of organic compounds is indicated by means of a quick-look technique developed by the petroleum industry. An apparent formation-fluid resistivity is calculated from Archie's law for a 100-percent water saturation. Constants for Archie's law are readily available in the literature.
Organic compounds in the formation can be calculated independently by using the Petra module in T-log by Terrascience. The organic compound becomes an additional component solved for in a multi component, mathematical matrix equation. The component used for this solution was ethane. Physical parameters are readily available for this compound. The output from this calculation shows percentages of sandstone, shale, effective porosity, and organic compound. This calculation required the use of the gamma ray log, the density log, the sonic log, and the resistivity log.
Formation-fluid resistivities for this calculation were obtained from the SP (self-potential) log. These resistivities were checked against those obtained from the quick-look techniques just explained. Other parameters for the multi component calculation were obtained from the literature and/or picked from the geophysical borehole logs.
The results from the quick-look and multi component solutions compare favorably. The results of these calculations are intended to aid in the targeting possible zones of TCE contamination. These mathematically calculated concentrations are rather low. Therefore, they should be thought of not as actual concentrations, but rather as indicative of possible "target" zones for organic pollutants.
Figure 3 show an example of a plot of these potential target zones. These figures also show the gamma ray log (corrected for steel casing GRCC), caliper, density, sonic transit time (filtered and unfiltered), sonic variable density, subunit marker identifications, shale-sandstone-porosity percentages, and suggested zones for seating packers for isolating potential target zones for future testing. These zones were selected to isolate potential TCE target zones and are determined from the sonic variable density to be areas of low fracture frequency. The caliper log was also used in determining the packer seats to assure uniformity of the well diameter in the seating area.
Analysis of the fracture data obtained from the borehole
televiewer (BHTV) is shown on Figure 4. Two figures for each well
comprise the fracture analysis. The first figure (Figure 5) for each
well presents fracture frequency and fracture aperture versus depth.
The second figure (Figure 6) for each well presents a fracture
orientation plot, a plot of fracture plane poles, and Kamb contours
of the fracture data. Data for each well is separated into three
sets- "A" natural fractures, "B" natural fractures and Bedding
Planes. The A set of fractures were readily identified from the BHTV
data and generally had apertures greater than 0.30 inches. The B set
of fractures were tabulated during a second review of the data and
generally had apertures of less than 0.30 inches. These small
aperture features are on the lower boundary of the resolution of the
BHTV system and should be assigned a lower confidence when
considering the fracture regime at the Butz Landfill Site. Although
bedding planes are commonly difficult to resolve with the BHTV
system, some bedding orientations were tabulated.
PERMEABILITYLog derived permeabilities are simply based on a porosity-permeability transform. A host of approaches are available to the log analyst. For this study both the Timur relationship and that developed at the Bureau of Reclamation were calculated. The Bureau relationship was developed to relate pump tests directly to log parameters. Using porosities derived from Archie's law allows the additional refinement of adressing different lithologies by means of proportionality constant and cementation exponent.
The Timur relationship is one of the more popular approaches where core data are not available and primary porosity is the dominant porosity. The results of the Bureau relationship yields an excellent fit to the permeability available from a slug test at Butz.
Log derived permeabilities can be adjusted by changing the coefficients in the mathematical log response equation to match pump tests if they are available. The advantage of log derived permeabilities is the resolution and capability of distinguishing aquifers and aquicludes with the same resolution obtainable from borehole geophysical logs. This allows refinements of choosing intervals for packer pump testing. Permeable zones(including fractures) can be defined and mapped with confidence.
Calculated permeability and transmissivity for seven of the seventeen wells logged are displayed on figures 7a through 7c.
Borehole geophysical logging at the Butz Landfill Site provides detailed information on lithology, fracturing, porosity, and possible contamination target zones. The conceptual model applicable at the Butz Landfill is shown in
Figure 8. The log analysis supports this model. Stratigraphic correlation using natural gamma logs indicates bedding in bedrock beneath the site generally dips 16 degrees to the north-northwest. No major offsets due to faulting are interpreted within the site area. Minor offsets may occur but cannot be determined due to distance between wells. Calculated permeability and transmissivity provide a first look method for estimating groundwater flow and contaminant movement. Further analysis of permeability and transmissivity for all wells geophysically logged may provide a hydraulic transmissivity contour map for bedrock.
Barton, C.A., L. Tesler and M.D. Zoback, 1991, Interactive analysis of borehole televiewer data, in Automated Pattern Analysis in Petroleum Exploration, edited by I.P.a.S. Sengupta, Springer Verlag, New York.
Hallenburg, J.K., 1984, Geophysical Logging for Mineral and Engineering Applications, pp. 254 PennWell Publishing Company, Oklahoma, 1984.
Parker, Beth L., R. W. Gillham, and J. A. Cherry, 1994, Diffusive disappearance of immiscible-phase organic liquids in fractured geologic media: Groundwater, Vol.32, No.5, pp. 805-820.
Schimschal, U., 1993, Density Adjustments in Air-Fill Boreholes in Volcanic Rocks at Yucca Mountain, Nevada: The Log Analyst, v. 34, # 4, p. 47-53.
Schimschal, U., 1981, "The relationship of geophysical
measurements to hydraulic conductivity at the Brantley damsite, New
Mexico", Geoexploration, v.19, p. 115-125, Elsevier Scientific
Publishing Company, Amsterdam.